[gmx-users] Beginner's problem: g_rdf functions and -com

[Henry -]

Dear Gromacs users,

This is my first time writing to this mailing list. I am writing in order to inquire about something which is after reading the manual quite extensively still unclear for me. I will explain my problem in detail, including details about my simulation. I hope you do not get discouraged by the length of this inquire.  So here goes: 

I have ran series of simulations with 2500 molecules where I have 10 Ethyl Acetate molecules and from 0 to 2490 molecules of Acetonitrile. The rest of the box is filled with, if any room available, water (SPC/E). Now I want to generate RDF-s around the Ethyl Acetate molecule in order to see the distribution of solvent and water.

My first question is: is it appropriate to calculate distribution when you have more than one reference atom (in this case I have 10 EtAc) or should I run new simulations where I exclusively use only one Ethyl Acetate molecule? Does GROMACS takes the distribution around every Ethyl Acetate molecule and averages it? 
Secondly, which one would be more correct, to use -rdf mol_com or not? If I am telling GROMACS to center around mass, does it center around each Ethyl Acetate and averages or does it take center-of-mass of 10 particles scattered around the simulation box and calculates distribution functions respect to that? Manual tells us that the default is "around a (set of)
particle(s)", how is that different from -com or -cog? Does default mode takes the reference particle as a sphere and calculates distribution of other species around it and -com takes the center-of-mass of the selected reference particle and does operations with that? 

I hope you understood my explanation and I hope I did not go too in detail and drop the readers' interest. I hope anybody has  time to answer this since it is quite important for my thesis.

Thank you in advance,

Yours faithfully,
Henry V